Member-only story
ML Algorithm cheatsheet for Interview
A practical refresher on the machine-learning algorithms I want to understand before walking into an interview
In this article, I am going to note down a summary of common machine learning(ML) algorithms I prefer take a quick look before any interview involving applied machine learning skills. I will not talk about the ‘continuous improvement’ part in this article.
Machine learning algorithms can look like a long list of unrelated methods: Linear Regression, Random Forest, SVM, XGBoost, PCA, neural networks, K-Means.
They become much easier to remember when I organize them by the problem they solve.
Supervised learning means I have predictors X and a known outcome y. I want to learn the relationship between them.
Examples: Linear Regression, Logistic Regression, Decision Trees, Random Forest, SVM, XGBoost, KNN and Naive Bayes.
Unsupervised learning means there is no outcome y. I am trying to discover structure in X.
Examples: K-Means, PCA, t-SNE and UMAP.
Time-series methods explicitly deal with observations ordered in time.
Examples: ARIMA and Exponential Smoothing.
